Our efforts to understand the structures of molecules relevant to immune recognition have, in the past year, focused on three prototype molecules: MHC class I molecules, T cell receptors, and natural killer (NK) receptors. These efforts are based on our laboratory?s ability to engineer, refold, purify and characterize in various binding studies the relevant molecules. With respect to the MHC class I molecules, a molecule that we have been particularly interested in for a number of years is the MHC-I molecule, H-2Dd. This molecule is of unique interest in that it not only binds specific T cell receptors, but it also binds the murine natural killer cell receptor, Ly-49A. The binding to T cell receptors is of particular interest because it has been shown by J. Berzofsky and colleagues, in collaboration with our laboratory, that MHC/peptide complexes formed of H-2Dd and P18-I10, can be recognized not only by CTL raised in the H-2d haplotype, but also by CTL raised in H-2p or H-2q. In addition, peptide H-2Dd complexes serve as the ligand on Ly-49A+ NK cells through which an inhibitory signal is generated. Thus, it has been of particular interest to determine the x-ray crystallographic structure of H-2Dd complexed with the HIV P18- I10, and to obtain structural information on the cognate ab TCR, alone and in complex with the MHC/peptide complex. We have engineered soluble versions of H-2Dd and a TCR that binds H-2Dd/P18-I10 complexes, expressed these proteins in bacteria and refolded and purified them to homogeneity from solubilized inclusion bodies. The interactions between these molecules have been examined by surface plasmon resonance and by analytical equilibrium ultracentrifugation. In collaboration with Dr. Roy Mariuzza?s group at CARB, University of Maryland, we have determined the crystal structure of the H-2Dd/P18-I10 complex to 3.2 ? resolution. Although there is overall similarity to other class I MHC proteins, there are several notable features that distinguish this structure. The bound peptide is in an unusual backbone configuration with the side chain of position 5 Arg directed deep into the binding site and the side chain of position 7 Phe almost totally exposed to solvent. This unique orientation of the peptide side chains was completely unexpected in terms of the predictions made by the interpretation of the results of functional experiments with individually substituted peptides. In retrospect, all of the functional data can be interpreted in the context of peptide side chains being either directly involved in T cell receptor interactions or indirectly involved. With respect to TCR structure, we have made significant progress in high resolution structural analysis of a TCR Va domain by both x-ray crystallography (in collaboration with Dr. E. Padlan, NIDDK) and by high resolution nuclear magnetic resonance (NMR) (in collaboration with Drs. J.-S. Shan and A. Bax, NIDDK). The x-ray crystal structure, solved at 2.5 ? resolution, reveals the Va domain with a basic immunoglobulin like fold. Good electron density is observed through the entire backbone of the molecule, with two molecules in the asymmetric unit of P212121 symmetry. The final model following rigid body refinement and crystallographic refinement, has a final Rcryst of 32.4% (Rfree22.4%) with 1794 non-hydrogen protein atoms and 24 solvent molecules. Of particular interest is that, unlike the previous Va4.2 previously solved by Mariuzza and colleagues, the Va2.6 fails to form a physiological dimer in the crystal. In comparison with other Va domains that have been solved, the major dissimilarities reside in CDR3 where length differences account for major differences in the loops. Other regions of dissimilarity include the first b strand and CDR1, where proline residues seem to be involved in determining unique Va structures. To extend and complement the structural studies of Va2.6, we have taken advantage of its high solubility to analyze this domain by multidimensional NMR, first under conditions that favor monomer formation, and then under the very high concentration conditions that allow some visualization of a dimer in solution. The three dimensional structure has been refined using structural constraints derived from NOE spectra, chemical shifts, J couplings, and dipolar couplings to achieve high accuracy and precision. Residues involved in the concentration dependent dimerization are in the process of being assigned. The analysis of 15N relaxation parameters of backbone amide sites reveals that all the secondary structure elements are non-mobile on the picosecond to nanosecond and on the millisecond time scale. A large number of slowly exchanging amide protons provides evidence for the stability of the Va core even on the scale of hours to days. Significant internal motions on the ps to ns time scale are detected for residues in CDR3, CDR1, and CDR2. With final refinement of the three dimensional structure by NMR, this will be compared to that of the Va in the crystal. Further studies allowing assignment of the specific residues involved in binding to the peptide/MHC complex in solution will be evaluated. Efforts to determine the structure of the natural killer cell receptors, Ly-49A, Ly-49C, and Ly-49G2 are now being undertaken as we have devised efficient methods of expressing these receptors in bacteria and refolding them in high yield. Characterization of the extracellular domain of Ly-49A in solution by analytical ultracentrifugation (P. Schuck, BSP, NIH) reveals that it is a monomer in equilibrium with a non-covalent dimer, described by a dimerization constant of ~10 ?M.